Liver X receptor (LXR) is a nuclear receptor that was cloned as an orphan receptor whose ligand and function were both unknown. Subsequent study reported that some oxysterols including 22-R-hydroxycholesterol act as a ligand for LXR (non-patent documents 1 to 3). LXR, together with retinoid X receptor (RXR) which is another nuclear receptor, forms a heterodimer to ligand-dependently control the transcription of a target gene.
As mammal LXR sub-types, two types of LXR genes (α and β) are known to exist. LXRα and LXRβ recognize the same sequence on a DNA and activate the transcription of a neighboring target gene. However, the expression-distributions of the two genes differ greatly. LXRα is specifically expressed on cholesterol metabolism-related tissues such as the liver, small intestines, or adipose tissues, whereas LXRβ is expressed ubiquitously on almost all tissues that have been examined (non-patent documents 4 and 5).
Many of the group of genes identified as target genes of LXRs are genes (ApoE, CETP, and LPL) related to a reverse cholesterol transport (RCT), including ABC transporters (ABCA1, ABCG1, ABCG5, and ABCG8). Therefore, it is expected that the activation of LXRs elevates the expression of these genes and activates reverse cholesterol transport pathways, thereby increases cholesterol efflux from the periphery and then increases HDL cholesterols and also lowers cholesterol content at an arteriosclerosis-affected region (non-patent document 6).
Further, LXRs are reported to play an important role via NF-κB suppression, in the expression control of inflammatory mediators such as NO-synthase, cyclooxygenase-2 (COX-2), and interleukin-6 (IL-6) (non-patent document 7). It is well known that the inflammation is very important at an arteriosclerosis-affected region, and it is expected that LXR ligands or LXR agonists will prevent arteriosclerosis exacerbation due to the expression of macrophage-inflammatory mediators at the affected region (non-patent documents 6 and 8).
Further, LXR α- and LXR β-deficient mice fed on high-cholesterol diet have been reported to show symptoms such as fatty liver and elevated LDL-cholesterol level as well as reduced HDL-cholesterol level in the blood as compared to the case of normal mice fed on high-cholesterol diet (non-patent documents 9 and 10). More specifically, it is strongly suggested that LXRs play an important role in cholesterol metabolism. Moreover, by analyzing the symptoms of arteriosclerosis mouse models having normal LXRα and LXRβ functions in the liver, small intestines and the like but lacking LXRα and LXR in macrophages, it has been revealed that LXRα and LXRβ activities in macrophages strongly affect the incidence of arteriosclerosis (non-patent document 11). Therefore, the activation of reverse cholesterol transport through the LXR activation especially in macrophages is considered to be important for the treatment of arteriosclerosis.
As for the applications, LXR regulators or LXR agonists disclosed in the prior art documents are reported to have been applied to diseases such as hypercholesterolemia and atherosclerosis (patent documents 1 and 2). Further, LDL-receptor-deficient mice loaded with high-fat food, and administered with LXR ligand, have been reported to show an elevated HDL cholesterol level, lowered VLDL and LDL cholesterol levels, and reduced area of arteriosclerosis-affected region (non-patent document 12).
Further, LXR ligands or LXR agonists are expected to control sugar metabolism in the liver and adipose tissues, and thus to improve diabetes (non-patent documents 6 and 8). Recently, it has been reported that an administration of LXR agonist improved insulin sensitivity and blood glucose level in diabetes animal models (non-patent documents 13 and 14). Moreover, it is indicated as a potential therapeutic drug for Alzheimer's disease, inflammatory diseases, or skin diseases (non-patent document 15).
LXR agonists, however, are reported to increase LDL cholesterol in animal species having cholesteryl ester transfer proteins (CETP) (non-patent document 16). Further, in animal experiments, it has been observed that LXR activation in the liver by the LXR agonist administration enhances fatty-acid and triglyceride syntheses through the transcriptional activation of enzymes that are important for fatty-acid synthesis, for example, fatty-acid synthase (FAS) or stearyl-CoA fatty-acid desaturase (SCD-1) (non-patent document 17). Meanwhile, nothing is disclosed in the prior art documents on LXR α/β selectivity in relation to the disclosed LXR regulators, LXR ligands, LXR agonists and the like.
Therefore, there have been demands for an ideal synthetic LXR-binding compound without a dyslipidemia-exacerbating effect which acts through an elevated fatty-acid and triglyceride syntheses, while maintaining the agonist activity for reverse cholesterol transport activation by ABC transporters and for increased cholesterol-efflux from macrophages. As one approach to solve the problem, a compound that selectively activates LXR is considered to have an ideal profile that is expected to suppress the activation of LXRβ highly expressed on the liver, as compared to the LXR regulators disclosed in the prior art documents, and to suppress the concerned side-effects of fatty-acid and triglyceride synthesis elevations (non-patent documents 6, 8, 15, 18, and 19). However, because ligand-binding sites of LXRα and LXRβ are highly homologous, it is considered that the creation of a compound that acts differently on LXRα and LXRβ is not easy.
In fact, compounds having an LXR-agonist effect have been reported, such as a benzofuran-5-acetic acid derivative (patent document 3), 2-aminoquinazolin-4-one derivative (patent document 4), tetrahydroquinoline derivative (non-patent document 5), tetrahydrocarbazol derivative (patent document 6), isoquinoline derivative (patent document 7), and naphthalene derivative (patent document 8), GW3965 which is an aromatic aminoalcohol derivative (Example 16 described in patent document 9), and T0901317 which is a benzenesulfonamide derivative (Example 12 described in patent document 10), but no agonist with high LXRβ selectivity has been reported to date and a compound with high LXRβ selectivity has been awaited.
Meanwhile, an LXR agonist having a quinoline skeleton has been reported (patent document 11, non-patent documents 20 to 22). For example, WAY-254011 (compound 4 of non-patent document 22) which is a quinoline derivative has been reported to have LXRβ-selective binding affinity (α/β ratio is 1 to 5). Non-patent document 22 further reports on a compound showing an α/β ratio of up to 1 to 50 in terms of binding-affinity. However, as for an agonist effect which was measured by Gal 4 transactivation activity, the highest selectivity confirmed was an α/β ratio of merely up to about 1 to 2.7. This shows that the effect of the compound on LXR for expressing the target gene is weak despite the selective binding of the compound to LXRβ. Therefore, there are still strong demands for a compound having an effect of expressing a target gene in an LXRβ selective manner.    [Patent Document 1] Published Japanese translation of PCT international publication No. 2002-539155    [Patent Document 2] Published Japanese translation of PCT international publication No. 2004-509161    [Patent Document 3] WO2003/82192    [Patent Document 4] WO2004/24161    [Patent Document 5] WO2004/72046    [Patent Document 6] U.S. Patent publication No. 2005/215577    [Patent Document 7] WO2004/58717    [Patent Document 8] WO2005/23188    [Patent Document 9] WO2002/24632    [Patent Document 10] WO2000/54759    [Patent Document 11] WO2005/58834    [Non-patent Document 1] Janowski et al., Nature, 383, pp. 728-731, 1996    [Non-patent Document 2] Lehmann et al., J. Biol. Chem., 272, pp. 3137-3140, 1997    [Non-patent Document 3] Fu et al., J. Biol. Chem., 276, pp. 38378-38387, 2001    [Non-patent Document 4] Auboeuf et al., Diabetes, 46, pp. 1319-1327, 1997    [Non-patent Document 5] Lu et al., J. Biol. Chem., 276, pp. 37735-37738, 2001    [Non-patent Document 6] Zelcer et al., J. Clin. Invest, 116, pp. 607-614, 2006    [Non-patent Document 7]: Joseph et al., Nat. Med., 9, pp. 213-219, 2003    [Non-patent Document 8] Geyeregger et al., Cell. Mol. Life. Sci. 63, pp. 524-539, 2006    [Non-patent Document 9] Peet et al., Cell, 93, pp. 693-704, 1998    [Non-patent Document 10] Alberti et al., J. Clin. Invest., 107, pp. 565-573, 2001    [Non-patent Document 11] Tangirala et al., Proc. Natl. Acad. Sci. USA, 99, pp. 11896-11901, 2002    [Non-patent Document 12] Terasaka et al., FEBS Lett., 536, pp. 6-11, 2003    [Non-patent Document 13] Cao et al., J. Biol. Chem., 278, pp. 1131-1136, 2003    [Non-patent Document 14] Laffitte et al., Proc. Natl. Acad. Sci. USA, 100, pp. 5419-5424, 2003    [Non-patent Document 15] Lala et al., Curr. Opin. Investig. Drugs, 6, pp. 934-943, 2005    [Non-patent Document 16]: Groot et al., J. Lipid Res., 46, pp. 2182-2191, 2005    [Non-patent Document 17] Schultz et al., Genes Dev., 14, pp. 2831-2838, 2000    [Non-patent Document 18]: Lund et al., Arterioscler. Thromb. Vasc. Biol., 23, pp. 1169-1177, 2003    [Non-patent Document 19] Bradley et al., Drug Discov. Today Ther. Strateg. 2, pp. 97-103, 2005    [Non-patent Document 20] Hu et al., J. Med. Chem., 49, pp. 6151-6154, 2006    [Non-patent Document 21] Hu et al., Bioorg. Med. Chem., 15, pp. 3321-3333, 2007    [Non-patent Document 22] Hu et al., Bioorg. Med. Chem. Lett., 18, pp. 54-59, 2008